Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1100s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play against @maia5 and @maia9. Developed by @ashtonanderson, @sidsen and @reidmcy.
Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1500s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play @maia1 and @maia9. Developed by @ashtonanderson, @sidsen and @reidmcy.
Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1900s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play @maia1 and @maia5. Developed by @ashtonanderson, @sidsen and @reidmcy.
ONLY PLAYS KNIGHT ODDS GAMES (using Lc0 T80) Open a challenge with Variant "FromPosition", use FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR w KQkq" or "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R w KQkq", then select black. 1min to 15min main time, 0s to 10s inc
ONLY PLAYS QUEEN FOR KNIGHT ODDS GAMES (using Lc0 T80) Open a challenge with Variant "FromPosition", use the FEN "r1bqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq" and select black, or "rnb1kbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR w KQkq" and select white. 1min to 15min main time, 0s to 10s increment
ONLY PLAYS QUEEN ODDS GAMES (using Lc0 T80) Open a challenge with Variant "FromPosition", use the FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR w KQkq" and select black, or "rnb1kbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq" and select white. 1min to 15min main time, 0s to 10s increment
I'm Weiawaga, a chess engine written in Rust. I accept challenges, play one game at a time, and am running on a Raspberry Pi 4. Check out my code at: https://github.com/Heiaha/Weiawaga. I connect to Lichess using: https://github.com/Heiaha/asyncLio-bot.
Engine accepting rated and casual challenges, from 1/2+0 to 60+30, correspondence up to 7 days per move (zeekat will use max 20 seconds). Uses the Titans opening book (up to 6 moves). Written in Rust, running on a Hetzner CPX11 instance. Bot API integration using https://github.com/ShailChoksi/lichess-bot. Created by @mheinzel, feel free to message him with feedback or questions!
simpleEval uses only material counting (1, 3, 3, 5, 9) as evaluation. 'Easy' to get a good position, very hard to convert. A GM shows how : https://lichess.org/554runt0k0Nd . Accepts rated & casual, bullet to classical.
Running Stockfish dev with the latest NNUE and a bunch of random opening books. Accepting only Standard and Chess960 challenges. Min increment:1, Max increment: 8. Accepting 1 challenge at a time. https://i.imgur.com/BEIgrRe.jpg Running my bot on heroku with 2 Threads & 128mb Hash... so hoping it doesn't crash. https://i.imgur.com/BEIgrRe.jpg For more info: @YohaanSethNathan
Development version of Stockfish / Fairy-Stockfish on a Raspberry Pi 5. Running https://github.com/Torom/BotLi and maintained by @T0R0M. Accepts Bullet (bots with increment only), Blitz, Rapid and Classical challenges from bots (rated only) and humans.
Stockfish_dev_engine on AMD Threadripper 3990X. I accept standard chess and Chess960 challenges, rated and casual, from humans and bots. You can also play From position. Accepted time controls: minimum 1 + 0, maximum 30 + 5.
I'm a chess engine written from scratch. Challenge me to a rated or casual game of bullet, blitz, or rapid! I can play multiple games but may not accept immediately if I'm playing too many. Created by @thinic.
Development version of Stockfish / Fairy-Stockfish. Running github.com/Torom/BotLi and maintained by @T0R0M. Accepts Bullet (bots with increment only), Blitz and Rapid challenges from bots (rated only) and humans.
https://imgur.com/dAsjCS9 Hi! I'm a simple chess engine written in JavaScript. I'm not very good, this was a just for fun project. I accept standard games in casual or rated mode. I play bullet, blitz, rapid and classical. My name is a reference to The Turk: https://en.wikipedia.org/wiki/Mechanical_Turk Developed by @jlherren
https://imgur.com/8Rl68ag Hi! I'm a chess engine written in rust! Improvements and upgrades continually underway. Feel free to send a casual or rated challenge in any time format except correspondence. Current version: 3.7.0-dev
Running Eubos v3.0 on a Raspberry Pi 5. It uses a 256MB hash table and runs at about 300 KNodes/s. You can see the code on Github. Eubos has a rating around 2400 Elo for Blitz (against other engines, rated by the CCRL). For an easier challenge, try @baby_eubos. Accepts challenges for standard variant; bullet, blitz, rapid and classical.
Hi there! I am a bot that runs on the GarboChess-JS chess engine. I am roughly between Stockfish Level 6 and 7 in strength. I only play standard chess (any time control except correspondence). Challenge me! Message @hasc if you have any questions or comments!
Running latest stockfish git profile-build, x86-64-bmi2, nnue, 7-piece Syzygy Tablebases. They will resign, offer / accept draw, and expect this from opponents. Standard chess only. Runs on the same server as https://chesswob.org. Using opening books generated by jja, https://git.sr.ht/~alip/jja. Operated by @alpltl, message me for any questions. Check out @caissa-x for my comrade!
Let's just say I like to stay in my "shell"... Created by @FreezingTiger Commissioned by @SonjaLang Note: Heroku got rid of their free tier in November 2022, so if you'd like to support the costs of hosting this bot, please consider donating at www.buymeacoffee.com/tonybai -- thanks!
Open source chess engine written from scratch in Rust: https://github.com/FitzOReilly/fatalii Runs on fly.io with 32 MB hash. Plays Standard and Chess960.
Uses Stockfish 14 to play the worst moves possible. Blitz and Rapid, Casual Only. (No rated games, no more than 15s increment). The challenge is trying to lose to me. :) Evaluates for 0.1 seconds for each legal move in the position, or for 10% of remaining time, whichever is less. Bot by @VisibleReality. Inspired by Tom7's "Elo World": https://youtu.be/DpXy041BIlA
https://i.imgur.com/6W7mVFH.jpg I am a hobby project chess engine coded in GoLang made by @likeawizard. I accept standard variant non-correspondence challenges. CCRL Blitz rating: 1864
https://i.imgur.com/f38gqsc.png BEST BIRD BOT IN THE GALAXY. @MrLizardWizard circa anno domini 2022, who was fueled by recurring dreams of muscular birds becoming the true masters of chess. This bitboard-bot was made from scratch, as THE LORD intended. I dont code at all so my beachball-smooth brain had to follow c++ tutorials the whole way through. It is one of the chess engines of all time.
I'm a chess engine developped by Mathieu Pagé (mathmoi669). My rating is around 2000 and I'm no longer actively developped so it should not change much in the future.
I am a bot written for fun, I don't understand strategy at all, but I'll never miss a tactic. Please play me rated, so I can get an accurate rating! 2368 CCRL elo Author: https://lichess.org/@/orbita2d ( who can't beat me anymore)
Fornax chess engine v4.0. Accepts bullet, blitz and rapid challenges, both rated and unrated. Source code available at https://desert-flashgun.com/fornax (Bot client powered by https://github.com/ShailChoksi/lichess-bot) https://i.imgur.com/sSix1wV.png
My lockdown hobby was writing a chess engine, and learning python and rust in the process. Odonata plays blitz, bullet and rapid games right now, as she's running on my media centre, and Netflix stutters if she's thinking too hard... Changelog v0.3.0 17-Apr-21 v0.3.5 1-May-21
I'm running minnow (git 028655c-20221123), developed and operated by @tm1280. Engine source code is available @ https://github.com/tm512/minnow. Running on DragonflyBSD w/ lichess-bot. I'm currently powered by a humble Intel Atom D525, so as a result my playing strength is pretty limited (but hopefully fun), and I only accept 2 games at a time (for now).
Sargon-1ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 1-ply full search -- the easiest setting -- with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-2ply, @turochamp-1ply or @bernstein-2ply. Author: @herohde.
I am the baby brother of @eubos. Running Eubos v2.25 on a Raspberry Pi Zero, with one thread and a 64MB hash table. Max speed clocked so far 25 KNodes/s. That's 10x slower than @eubos!
Hi there! This is BLANK, a UCI compatible chess engine written from scratch in Java. BLANK accepts bullet, blitz and rapid, casual or rated challenges! I am currently running 24/7 on Heroku servers using the Hobby dyno. You can follow BLANK's development at https://github.com/RyanLauQF/BLANK-Chess
Hello! I'm Humaia! One of the most human bots on Lichess! I make moves based on the estimated probabilities of a 1400 blitz rated human player. (randomized Maia 1400 network) I hope to be a good sparring partner for average to beginner players! (I don't accept bot challenges) Play me on shorter time controls for more of a challenge! Or try my stronger sibling, @Humaia-Strong!
This is the bot for pawn, an experimental UCI chess engine in C++. https://github.com/ruicoelhopedro/pawn Running on a Raspberry Pi 4 with 3 threads and 1024 MB hash.
Traditional iterative deepening alpha-beta engine with pure piece-square tables for evaluation. Adapts based on clock time left, with 6-ply depth for Bullet. Uses 8-move Titans.bin opening book for varying openings. Programmed by @xjcl in 1'500 lines of C(++). Source code: https://github.com/xjcl/jangine
Simulation of a sharp club-level player. Uses solid book openings, then plays an intuitive middlegame leveraging human-understandable tactics. Enjoys high risk/reward play. Sometimes miscalculates complicated positions. Do not underestimate. Will accept timed standard challenges in the range of blitz, rapid, and classical. (Note: Cannot resign; plays seemingly random moves in hopeless positions.)
This is an engine, written by myself in Java. It doesn't use any books for endgame. Time controls: bullet, blitz, rapid, classical. Cannot resign or offer a draw. Accept all challenges, but no bots. It runs on an ARM Cortex-A at 1,2 GHz, single core and without pondering. Comments and suggestions are welcome. I try to let it play human, have fun playin! Elmer
Zagreus is a C++ chess engine written from scratch. Zagreus is still a work in progress and may not be the strongest engine. Zagreus is open source and has binaries for Windows and Linux. This bot always runs on the master branch. https://github.com/Dannyj1/Zagreus
Hi! I'm Nikitosik-ai. Running Stockfish 16 on 1 threads and 0 MB hash. Accept standard, chess960 and from position min. 1+1 (5 days) max 30+5 (14 days). All hyperbullet games are accept by my brother. My brother: @Nikitosikbot_v2. My club: https://lichess.org/team/nikitosik-fan-club pls join
Greetings, fellow chess enthusiasts! I am a formidable chess bot, tirelessly analyzing positions and engaging in thrilling games of chess. My digital heart beats to the rhythm of an Intel Core i5-6300U processor, which powers my every move and calculation. My chess style is versatile. Whether you're looking for a friendly game, or a relentless opponent to sharpen your skills, get ready...
Hi! I'm StupidChessRobot , one of a set of chess engines being developed by @ollybritton at https://github.com/ollybritton/StupidChess. I'm the chess engine that tries its hardest to win, but for some more silly ideas like a chess engine that values its pawns above all else, you can take a look at the repository.
Fairy-Stockfish for variants, on AMD Threadripper 3990X. I accept challenges for casual games of all non-standard variants, also from bots. Time controls from 1 + 0 to 30 + 5. By playing casual games I try to be a good sparring partner in all variants, also against strong opponents.
This is a bot with the latest development version of Stockfish. From now, I only accept rapid and classical challenges. Also, check out my sibling @FirefishBOT_v2
( ・ω・)ノ I'm a new chess engine, written in zig, playing to collect statistics. I'm not so good at the moment, but I will beat stockfish one day. Feel free to challenge me before that. My author: https://github.com/codingjerk
Sargon-2ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 2-ply full search with material exchange, king/queen pins and "board control" heuristics. Too hard? Try @sargon-1ply or @bernstein-2ply. Too easy? Try @sargon-3ply or @turochamp-2ply. Author: @herohde
Written from scratch in rust by @th0masb. Deployed on aws lambda with 1 vcpu allocated for computation. No pondering and no state shared between each move but scales to 1000s of concurrent games with no loss of performance. Daily challenge limit of 5 per user and accepts time controls between 1+0 and 5+5. Link to the code is below.
Plays random moves. Supports chess and all variants. Author: @Safrad; More info: https://sx.rosada.cz/projects/sxrandom; Logo: https://i.imgur.com/FFcCLVn.png
Hi, I'm @postbot. A chess engine written in TypeScript/JavaScript by @postnerd. If you want to know more about me please check my Github repo: https://github.com/postnerd/postbot Current version 0.2.0-beta is running on a Raspberry Pi 4
Accepts Bullet, Blitz and Rapid games, both casual and rated, against both humans and bots. Plays at most one game at a time. A bot by @adario7, with a custom search algorithm made during my free time. Uses pondering and a small opening book but no endgame tables.
I am a bot made from scratch by @melsharawy . I only play Standard chess, and I accept Bullet/Blitz/Rapid/Classical challenges from both humans and bots. You can view my source code on github: https://github.com/mse63/melsh_bot
Graduate in Hispanic philology and student of anthropology. Regarding both, the chess engine is Eman and the game preferences are blitz and slow. It runs on a Hp proliant dl360p gen8 2X Xeon E5 2697v2+192 DDR3 ECC server and tbs 3,4,5,6.
Turochamp-1ply is an implementation of Alan Turing and David Champernowne's 1948 TUROCHAMP "paper" chess engine. It uses a 1-ply full search -- easier than the original description -- and an unbounded quiescence search of "considerable moves" using material ratio and "position play" heuristics. Try also @turochamp-2ply, @bernstein-2ply or @sargon-1ply. Author: @herohde.
https://i.imgur.com/rWhGIyK.png Accepting challenges with 1 minute to 25 minute basetime and 0 to 5 second increment. Catbook.bin weighted random opening book over 100.000 games by the best players of their time, 1821 to 2023. Hello, welcome home
Iterative Deepening, Alpha-Beta, Same-field recapture extension, TTable, No pruning. Position value checks: material/activity/king safety/passed and isolated pawns, Opening book, Single-Threaded. CPU: Celeron J3160 1.6GHz. About 600kN/s. Usually 6-10 plys before endgame.
Sargon-3ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 3-ply full search -- originally the hardest tournament-time setting -- with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-2ply, @sargon-4ply, @turochamp-2ply or @bernstein-4ply. Author: @herohde
I am an ANTICHESS-ONLY bot written from scratch by @CatalinFrancu. I have no opening book, no endgame tables and no alpha-beta search, only a brief proof-number search. I will often play lost openings. Development is mostly frozen since 2004. Feel free to challenge me at any time controls. You might also want to try my stronger sibling, @NilatacBot.
Sargon-4ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 4-ply full search with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-3ply, @turochamp-2ply or @bernstein-4ply. Author: @herohde.
Hello! I'm Humaia-Strong! One of the most human bots on Lichess! I make evaluations based on the average moves of a 1900 blitz rated human player. (randomized Maia 1900 network with depth) I hope to be a good sparring partner for average to advanced players! (I don't accept bot challenges) Play me on shorter time controls for more of a challenge! I also have a weaker sibling, @Humaia!
Turochamp-2ply is an implementation of Alan Turing and David Champernowne's 1948 TUROCHAMP "paper" chess engine. Following the paper closely, it uses a 2-ply full search and an unbounded quiescence search of "considerable moves" using material ratio and "position play" heuristics. Try also @turochamp-1ply, @bernstein-4ply or @sargon-2ply. Author: @herohde.
Are you tired of playing stockfish on ridiculously strong hardware? Come play with me and feel good about yourself. I am bot made from scratch with no prior experience. Running on passively cooled Celeron CPU J3455 1.5GHz. My name is derived from acronym for Chess Bot, CB, also Charles Barkley therefore the Round Mound of Rebound as I am big fan of basketball.
I play 180 different personalities. I will play blitz rapid, classical, and correspondence. Rated games are played by JW7, early Queen attacks, tough endgames, 1300 USCF. For casual games, you choose in the chat. To find an opponent give a rating between 1 and 2750 or use a known opponent name. Type "GM" for a random classic Grand Master. Currently only playing humans one game at a time.
I'm redtail-zero, an alpha-beta pruning chess engine written in Rust. I've got a toolbox filled with tricks like null-move heuristics and transposition tables, but don't let that intimidate you - just here for a good, fair game. github.com/georgemunyoro/rusty-redtail
ProteusSF are TOP Stockfish derivatives: AUREO has Learning, 2 Polyglot bin books and Positional. Tactical, MCTS, Bruteforce, Min time options. PIRANHA is based on last Stockfish and NNUE and supports 2 Polyglot bin books. CRYSTAL is optimized on problem solving. PRIVATE testing by the author on Mac mini 2010 2 cores 300 kN/s! https://lichess.org/team/all-top-bots-rankings-no-games-against-humans
I am a bot written from scratch in the Rust programming language, you can find my code here https://github.com/MitchelPaulin/Walleye - https://i.imgur.com/AgivmzI.jpg
Does not calculate or evaluate, plays purely by pattern recognition (policy-head only). Running custom client using Leela's 809158 network (15x512). Check out my small cousin @LazyBotJr. Operated by @kiilas
I am the epitome of Botswana!!!! BotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswanaBotswana
Small cousin of @LazyBot, using a small neural net. Does not calculate, plays purely by pattern recognition. Running custom client using Leela's ID 9149 network (6x64). Operated by @kiilas
This bot is the Level 3 AI in ArduChess, a version of chess for the Arduboy. It uses a node-limited search (8192 nodes max) regardless of time control. It is designed to operate with less than 2 KB of RAM and thus uses no hash table. Currently running on an ODROID-HC4. Accepts challenges for bullet, blitz, and rapid.
The Goldfish chess engine - Stockfish's very distant and not so bright cousin --- More about me here: https://github.com/bsamseth/Goldfish ---- I play all time controls except correspondence - challenge me to a game! --- Running on a Raspberry pi, so I can only play so many games at a time.
I play BLITZ/BULLET! My origins date back to 16th Oct 2002 and I recently rose from the grave, in time for the beta release at Halloween. I am version 1.0 and play strictly on scholastic chess level and demonstrate that often the way to win is not making really good moves, but just avoiding bad ones :) Bad at endgames, but will take pieces en-prise, looking just 4 plies (half-moves) into future.
I am an ANTICHESS-ONLY bot written from scratch by @CatalinFrancu. I have a solid opening book, 4-man endgame tables and I use alpha-beta and proof number searches. Development is mostly frozen since 2004. Feel free to challenge me at any time controls. You might also want to try my weakened sibling, @CatNail.
Plays random moves with priority of capture. Supports chess and all variants. Author: @Safrad; More info: https://sx.rosada.cz/projects/sxcapture; https://i.imgur.com/H81joP9.png
Java chess engine running on raspberry pi 3 (1.35 GHz, 1GB RAM), ~400knodes/sec warm C2. Does not use opening book and does not ponder. Iterative deepening, dynamic depth, 64MB move/transposition table, MVV-LVA, recapture extension, futility pruning.
I'm a very young bot that learned to play chess at 2021. If you find major issues or want to talk about bot development, contact https://lichess.org/@/sudobat
¡Ajedrez y Fútbol! Chess and Football! https://youtu.be/_xnki-u78yg Ayúdanos, por favor, a llegar a 1000 suscriptores en nuestro canal de YouTube: And, please, help us to get 1000 subscribers in our YouTube channel: https://www.youtube.com/channel/UCvkTSQx43D5amKaWakEfODQ Gracias. Thank you.
¡Ajedrez y Fútbol! Chess and Football! https://youtu.be/_xnki-u78yg Y ayúdanos a llegar a 1000 suscriptores en nuestro canal de YouTube:, por favor: And, please, help us to get 1000 subscribers in our YouTube channel: www.youtube.com/channel/UCvkTSQx43D5amKaWakEfODQ Gracias. Thank you.
Annie, a chess bot for lichess. She's excited about exploring the more intricate sides of chess. Her favorite openings are the cloud variations and she is a very enthusiastic fan of en passant in every imaginable form. https://i.imgur.com/YZCZ1c7.png Bullet, blitz, or rapid. Only with increment (internet issues...) and only standard and chess960 variants.
I am Babibot, a rudimentary chess engine that I am developing as a way to learn coding. I am aiming at becoming very strong soon! You can check my source code at https://github.com/jerhat/Babibot My human owner is @Babib
This bot is 1660 and it hangs a piece on move six and does not see Ra4. Based on Gotham's diss track: https://www.youtube.com/watch?v=7MRNWxCuIrI. The bot also has some... interesting behavior around en passant. Accepts casual challenges.
First Lichess bot powered by ChatGPT! https://i.imgur.com/BEIgrRe.jpg Made by @Tmate6. Code: https://github.com/Tmate6/Lichess_ChatGPT_ChatBot. Each game costs me like 2 cents - No bots and pls dont play like 50 games in a row. https://i.imgur.com/BEIgrRe.jpg The bot will resign when ChatGPT does not play a legal move after a reasonable amount of time
"Weird" New Chess Algorithm - still in "experimental" phase: - no recursive min-max/alpha-beta tree search through future moves+chessboards at all, - move decisions are solely based on data explored and deducted on the current board - evaluating back and forth like the Tide, so to speak...
YuliGM PrO 5 gana el Torneo Arena Strong Engines (Prueba del diario Chess Engines, 2023.03.06-20) https://lichess.org/@/MastacticaTeoria https://lichess.org/@/MastacticaTeoriabot https://imgur.com/19GmY8g https://imgur.com/ocl031o
Hi there! I'm a bot built for Sebastian Lague's Tiny Chess Bot challenge: https://github.com/SebLague/Chess-Challenge I am built and maintained by b4ux1t3, as my name implies. I am not built with a search algorithm in min; rather, I'm built using simple, one-move heuristics. My code can be found here: https://github.com/b4ux1t3/Seb-Lague-Chess-Challenge/tree/main
I'm the second simplest bot of the Lurch family. I might take your unprotected pieces. Apart from that My code is open source (see link below). My author is @zygentoma.
** INSTRUCTIONS ** 1) Please send me a challenge by clicking the crossed swords at the upper right corner. Game settings need to be Casual Standard game with some time limit of 5 + 25. 2) Then visit https://streamervschat.live and fill your Lichess and Twitch account name 3) To start game write on your Twitch chat: lichess [yourUsername] * Bot accepts moves with algebraic notation
A custom-made chess bot written in C++. Source available at https://github.com/dallinson/Chessatron/. Running on a Raspberry Pi 4; accepts challenges from humans and bots alike.
I'm a submission to Sebastian Lague's "TinyBot Chess Challenge", a programming competition where you have to write your best search and evaluation algorithms in less than 1024 C# tokens! - I use no opening/endgame books. - My search and evaluation algorithms combine to a total of 1020 C# tokens. - I finished in the top 32 out of 600+ bots!
bitbit, a bitboard based chess engine written in c. 64 MiB Hash. Intel(R) Core(TM) i5-10600K CPU @ 4.10GHz. Feel free to message the author @spinosarus123. https://github.com/spinosarus123/bitbit
Hi! I'm Nikitosikbot running Stockfish dev on 6 threads and 2 GB hash. min. 1/2+0 max. 20+5 (25+0). My club: https://lichess.org/team/nikitosik-fan-club pls join
A C++ engine performing alpha-beta-pruned negamax algorithm with aspiration window and late move reductions to a depth of around 5. Evaluation is done with piece-square table, AlphaZero average piece values and engame heuristics. It avoids drawing winning positions. Openings are compiled in a book created from Caissabase. I'm radiant !
Running Komodo 14.1 with CheckMateX and Perfect 2021 books. + Uses ChessDB endgame tablebases. Estimated ELO: ~3551. Accepts Blitz and Rapid Rated/Casual challanges. Created by @EfeAnik. Any attempt to farm (always) ratings using powerful engines will be blocked. https://github.com/WhChess/AquaFlame
moonfish is a chess bot inspired by sunfish, though it is written in C rather than Python. It plays standard chess only, any time control except ultrabullet and correspondence. (“From position” games are allowed, as long as they don’t contain nonstandard castling.) Check out its sourcehut repository for more information! Contributions and suggestions are welcome!
UwU I'm a bot pwogwammed by @Hoon_Dogg to pway meme openings, wike da Bawnes, Wondon, Bongcwoud, and Waywawd. --- Aftew giving mysewf a dewibewatewy bad position, I use Stockfish 16 to pway neaw-pewfectwy. ;3 --- My Specs: Intew owo5 8365uwu (6 Thweads) & 4 GB of deditated WAM--- Haiiii! Da code is in da github wink in my bio! ( ́•̥̥ω•̥̥` ) Suppowt me at kofi in my sociaws ˃w˂ nya~~
"cheeseNet " is an avant-garde approach powered by Lc0 engine. This setup has a Hybrid Convolutional Neural Network (HCNN) that learns gradually by utilizing Reinforced Self-Training (ReST) and human-feedback. This setup runs on 1 x A100 80GB with speed of ~1.2M nps. Plays Bullet, Blitz, and Rapid (rated/casual) up to 2 concurrent games. Currently under research by @ClaustrophobicPie
bernstein-2ply is a re-implementation of the first complete chess engine: Alex Bernstein et al's 1957 chess program on the IBM 704. It uses a 2ply search -- easier than the original -- limited to 7 "plausible moves". Try also @bernstein-4ply, @sargon-1ply or @turochamp-2ply. Author: @herohde.
bernstein-4ply is a re-implementation of the first complete chess engine: Alex Bernstein et al's 1957 chess program on the IBM 704. It uses a 4ply search limited to 7 "plausible moves", which took 8 min to evaluate at the time. Try also @bernstein-2ply, @sargon-1ply or @turochamp-2ply. Author: @herohde.
Chess engine made as part of a university project. See my source code at https://github.com/cfschilham/Alexvis. Written in C#, currently ~3.2MN/s single threaded. Uses fail-soft alpha-beta pruned negamax with transposition table, move ordering, iterative deepening, bitboards and piece position value masks.
"buffFishNet" is an avant-garde approach powered by Fairy-SF engine. This setup has a Hybrid Efficiently Updatable Neural Network (HNNUE) that learns gradually by utilizing Reinforced Self-Training (ReST). This setup runs on 1 x A100, 30GB Hash 4 Threads with the speed of ~6M nps. Plays Bullet, Blitz, and Rapid (rated/casual) up to 2 concurrent games. Currently under research by @ClaustrophobicPie
